初识 机器学习/深度学习

• 1. 参考教程
• 2. 服务器资源
• 3. 环境配置
  • 3.1. cudnn, CUDA 的区别
  • 3.2. 查询 CUDA 版本
  • 3.3. 查询版本
• 4. 训练集/验证集/测试集
  • 4.1. K 折交叉验证法
• 5. 过拟合
• 6. 环境
• 7. 数据集
  • 7.1. 图像标注
• 8. 数据预处理
• 9. 训练与预测
• 10. 验证
• 11. 预测结果导出
• 12. 常用函数

1. 参考教程

1. 吴恩达老师的深度学习课程：DeepLearning.AI: Start or Advance Your Career in AI: https://www.deeplearning.ai/
2. 李宏毅老师的深度学习课程
3. 《deep learning》Deep Learning（花书）: https://www.deeplearningbook.org/
   1. 中文版 exacity/deeplearningbook-chinese: https://github.com/exacity/deeplearningbook-chinese
4. 周志华老师的《机器学习》（西瓜书）
5. 《动手学深度学习》 — 动手学深度学习 2.0.0-beta0 documentation https://zh-v2.d2l.ai/index.html
6. 深度之眼官方账号的个人空间_哔哩哔哩_bilibili: https://space.bilibili.com/365093772

2. 服务器资源

1. SageMaker Studio Lab https://studiolab.sagemaker.aws/

3. 环境配置

NVIDIA cuDNN | NVIDIA Developer : https://developer.nvidia.com/cudnn
CUDA Zone | NVIDIA Developer : https://developer.nvidia.com/cuda-zone

3.1. cudnn, CUDA 的区别

显卡，显卡驱动,nvcc, cuda driver,cudatoolkit,cudnn到底是什么？ - 知乎 : https://zhuanlan.zhihu.com/p/91334380

1. cudnn 调用 CUDA 中的工具
2. Pytroch 中包含 cudnn 库

3.2. 查询 CUDA 版本

# Windows Linux 通用
nvcc --version

import torch
print(torch.__version__)

print(torch.version.cuda)
print(torch.backends.cudnn.version())

3.3. 查询版本

# 查询 opencv 版本
import cv2
cv2.__version__

4. 训练集/验证集/测试集

训练集：训练数据集，用于训练模型
验证集：用于对模型的少量调整(训练过程中多次调用)
测试集：用于评估模型的效果

通常三个数据集的数据量比例为：6:2:2

4.1. K 折交叉验证法

5. 过拟合

6. 环境

# This Python 3 environment comes with many helpful analytics libraries installed
# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python
# For example, here's several helpful packages to load

import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)

# Input data files are available in the read-only "../input/" directory
# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory

import os
for dirname, _, filenames in os.walk('/kaggle/input'):
    for filename in filenames:
        print(os.path.join(dirname, filename))

# 查看当前根目录
print(os.path.dirname(os.path.abspath("__file__")))

7. 数据集

7.1. 图像标注

1. labelme
2. [在线] Make Sense : https://www.makesense.ai/
   1. 暂未找到标记错误时如何撤销

8. 数据预处理

import pandas as pd
    
# Load data
melbourne_file_path = '../input/melbourne-housing-snapshot/melb_data.csv'
melbourne_data = pd.read_csv(melbourne_file_path) 
# Filter rows with missing values
melbourne_data = melbourne_data.dropna(axis=0)
# Choose target and features
y = melbourne_data.Price
melbourne_features = ['Rooms', 'Bathroom', 'Landsize', 'BuildingArea', 
                        'YearBuilt', 'Lattitude', 'Longtitude']
X = melbourne_data[melbourne_features]

from sklearn.model_selection import train_test_split

# split data into training and validation data, for both features and target
# The split is based on a random number generator. Supplying a numeric value to
# the random_state argument guarantees we get the same split every time we
# run this script.
train_X, val_X, train_y, val_y = train_test_split(X, y,random_state = 0)

9. 训练与预测

from sklearn.tree import DecisionTreeRegressor
# Specify Model
iowa_model = DecisionTreeRegressor(random_state=1)
# Fit Model
iowa_model.fit(train_X, train_y)

10. 验证

# Make validation predictions and calculate mean absolute error
val_predictions = iowa_model.predict(val_X)
val_mae = mean_absolute_error(val_predictions, val_y)
print("Validation MAE when not specifying max_leaf_nodes: {:,.0f}".format(val_mae))

11. 预测结果导出

output = pd.DataFrame({'Id': test_data.Id, 'SalePrice': test_preds})
output.to_csv('submission.csv', index=False)

12. 常用函数

# 导入数据

# 读取列标签
data.columns